Awareness and Knowledge of Clinical Practice Guidelines for CKD Among Internal Medicine Residents: A National Online Survey

Results 

A total of 651 respondents accessed the online survey. After excluding incomplete questionnaires and responses from 16 attending physicians and 9 chief residents, our final sample size consisted of 479 internal medicine residents from 75 residency programs who completed the questionnaire. Their PGY distribution was 166 PGY1 (34.7%), 187 PGY2 (39.0%), and 126 PGY3 (26.3%). The response rate could not be calculated because we did not know how many residents received the questionnaire from their program directors.

Of the residents, 204 (42.6%) were women and 272 (56.8%) had international training (medical school/residency/fellowship) (Table 1). The geographic distribution of the residency programs (n = 75) was 27 Northeast (36%), 23 Midwest (31%), 14 South (19%), 10 West (14%) and 1 Pacific (1%). Two hundred fifteen residents (44.7%) were from university-based programs and 270 (56.5%) were from programs with in-house nephrology fellowship programs. For management of CKD, 204 residents (42.6%) claimed to be using the KDOQI guidelines and 132 (27.6%) were not aware of such guidelines, whereas 160 residents (33.4%) believed the hypertension (Seventh Joint National Committee [JNC-7]) and diabetes (American Diabetes Association [ADA]) guidelines were adequate.

Table 1. Baseline Demographic Characteristics of Internal Medicine Residents in the Study

		PGY1 (n = 166)	PGY2 (n = 187)	PGY3 (n = 126)	P	All PGYs (n = 479)
	Women (%)	49.4	39.5	38.3	0.1	42.6
	International training (medical school/residency/fellowship) (%)	55.4	63.6	48.4	0.03	56.8
	University-based residency program (%)	44.6	41.2	50.0	0.3	44.7
	In-house nephrology fellowship program (%)	55.5	52.2	64.3	0.1	56.5
	Guidelines used to manage CKD
	KDOQI (%)	25.9	43.9	62.7	<0.001	42.6
	Not aware of any guidelines for CKD (%)	42.8	26.7	8.7	<0.001	27.6
	Hypertension (JNC-7) and diabetes (ADA) guidelines (%)	29.5	34.8	36.5	0.4	33.4

Abbreviations: ADA, American Diabetes Association; CKD, chronic kidney disease; JNC-7, Seventh Joint National Committee; KDOQI, Kidney Disease Outcomes Quality Initiative; PGY, postgraduate year.

The survey instrument had modest internal consistency (Cronbach α = 0.69). Cronbach α with deletion of 1 item ranged from 0.61 to 0.66, suggesting that the questions had similar influence on the total performance score.

Definition and Classification of CKD (section score = 2) 

A clinical vignette was given describing a patient with positive proteinuria tested twice in 3 months and an estimated glomerular filtration rate (GFR) of 76 mL/min/1.73 m2 (Table 2; estimated GFR in mL/min/1.73 m2 may be converted to mL/s/1.73 m2 by multiplying by 0.01667). Two hundred fifty-eight residents (53.9%) correctly recognized proteinuria, ie, kidney damage to indicate CKD, and 134 residents (27.9%) defined CKD as equivalent to GFR less than 60 mL/min/1.73 m2. A question was asked about the classification of CKD: “What is the level of GFR in stage III CKD?” (Table 2). Three hundred ten residents (64.7%) correctly answered “30-60 mL/min/1.73 m2.” The proportion of correct responses increased by PGY level (Table 2).

Table 2. Resident Awareness of Definition, Risk Factors, and Laboratory Evaluation of CKD

	A 49-year-old African American man is in your medicine clinic. He does not have hypertension or diabetes. His laboratory test results show serum creatinine level of 1.3 mg/dL, eGFR of 76 mL/min/1.73 m2, and urine dipstick result of 1+ protein. Additional urine studies show proteinuria (quantitative), repeated twice in 3 months.
		PGY1 (%) (n = 166)	PGY2 (%) (n = 187)	PGY3 (%) (n = 126)	P	All PGYs (%) (n = 479)
	According to the definition of CKD
	He has CKD because he has proteinuria tested twice	42.2	59.4	64.3	<0.001	53.9
	He needs to have GFR < 60 mL/min/1.73 m2 to have CKD	31.9	26.7	24.8	0.4	27.9
	He needs to have his urine tested again for proteinuria because 3 abnormal urine test results suggest CKD	9.0	8.0	7.7	0.9	8.7
	I don't know	16.9	5.9	3.2	<0.001	9.5
	Classification of CKD
	eGFR range of 30-60 mL/min/1.73 m2 for stage III CKD	51.8	69.5	74.6	<0.001	64.7
	Risk factors for CKD
	Age > 60 y	68.7	71.1	73.8	0.6	71.0
	African American/Hispanic	70.5	66.3	69.0	0.7	68.5
	Diabetes mellitus	98.8	98.9	99.2	0.9	99.0
	Hypertension	98.2	99.5	99.2	0.5	99.0
	Obesity	36.7	40.1	38.1	0.8	38.4
	Systemic lupus erythematosus	86.7	80.2	82.5	0.3	83.1
	Daily NSAID use	80.7	73.8	70.6	0.1	75.4
	Family history of CKD	84.3	71.1	73.0	0.009	76.2
	Laboratory evaluation of persons at increased risk of CKD
	Serum creatinine alone	15.7	10.7	9.5	0.2	12.1
	Serum creatinine to estimate GFR	86.7	86.6	88.1	0.9	87.1
	Urinalysis with microscopic analysis	74.7	77.5	80.2	0.5	77.2
	Urine dipsticks to estimate protein or albumin	49.4	39.0	45.2	0.1	44.3
	Random urine albumin or protein	27.7	19.8	19.0	0.1	22.3
	Random urine albumin-creatinine or protein-creatinine ratio	62.7	77.0	80.2	0.001	72.9

Note: Results are expressed as percentage of residents who chose an option. eGFR in mL/min/1.73 m2 may be converted to mL/s/1.73 m2 by multiplying by 0.01667; serum creatinine in mg/dL may be converted to μmol/L by multiplying by 88.4. Abbreviations: CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate; NSAID, nonsteroidal anti-inflammatory drugs; PGY, postgraduate year.

Clinical Action Plan for CKD (section score = 7) 

Residents were asked to specify the target blood pressure (BP) in a 52-year-old patient with diabetic CKD stage II and microalbuminuria (Table 3). Target BP less than 130/80 mm Hg was selected by 287 residents (60.1%). In addition, another 140 residents (29.2%) chose goal BP less than 125/75 mm Hg. Other selected interventions to slow the progression of CKD in this patient included use of angiotensin-converting enzyme (ACE) inhibitors or angiotensin receptor blockers (ARBs; 468 residents; 97.7%), aggressive glycemic control (436 residents; 91.0%), dietary salt restriction (301 residents; 62.8%), lipid control (337 residents; 70.4%), weight loss (372 residents; 77.7%), and smoking cessation (356 residents; 74.3%). In this section, the performance score was similar across all PGY levels (Table 3).

Table 3. Resident Awareness of Goal BP, Clinical Action Plan, and Complications of CKD

	A 52-year-old Caucasian woman with type 2 diabetes mellitus comes to you with BP of 148/92 mm Hg and the following laboratory test results: serum creatinine, 0.9 mg/dL; eGFR, 70 mL/min/1.73 m2; and urine study, microalbuminuria (urine albumin-creatinine ratio, 58 mg/g). What is the goal BP in this patient?
		PGY1 (%) (n = 166)	PGY2 (%) (n = 187)	PGY3 (%) (n = 126)	P	All PGYs (%) (n = 479)
	Goal BP < 140/90 mm Hg	1.8	0.4	0.0	0.06	0.8
	Goal BP < 135/85 mm Hg	4.4	3.4	4.0	0.9	3.8
	Goal BP < 135/80 mm Hg	5.6	6.4	6.3	0.9	6.1
	Goal BP < 130/80 mm Hg	56.9	60.4	63.5	0.5	60.1
	Goal BP < 125/75 mm Hg	31.3	29.4	26.2	0.6	29.2
	What is the clinical action plan for this patient?
	Start her on ACE/ARB therapy	95.2	99.5	98.4	0.02	97.7
	Dietary salt restriction < 2.4 g/d	62.0	66.3	58.7	0.4	62.8
	Lipid control	63.3	74.9	73.0	0.04	70.4
	Glycemic control	87.3	91.4	95.2	0.06	91.0
	Weight loss if obese	74.7	79.1	79.4	0.5	77.7
	Smoking cessation	71.7	74.3	77.8	0.5	74.3
	Which medications help reduce proteinuria independent of its effect on BP?
	ACE inhibitor/ARB	91.0	98.4	96.8	0.003	95.4
	Diuretics	3.0	4.3	1.6	0.4	3.1
	Nondihydropyridine calcium channel blocker	6.6	12.3	16.7	0.03	11.5
	Dihydropyridine calcium channel blocker	4.8	9.1	11.1	0.1	8.1
	β-Blockers	2.4	5.9	0.8	0.04	3.3
	Referral to nephrologist
	At eGFR < 30 mL/min/1.73 m2	86.1	88.2	96.0	0.02	89.6
	What are the potential complications of CKD for eGFR <60 mL/min/1.73 m2?
	Anemia	88.0	92.5	92.1	0.3	90.8
	Bone disease	76.5	84.5	85.7	0.07	82.0
	Coronary artery disease	39.8	62.6	58.7	<0.001	53.7
	Stroke	30.1	42.8	48.4	0.004	39.9
	Malnutrition	31.3	41.2	40.5	0.1	37.6
	Increased risk of such diabetic complications as retinopathy	36.1	44.9	40.5	0.2	40.7
	Medication complications (eg, acute renal failure)	75.3	76.5	81.7	0.4	77.5

Note: Results are expressed as percentage of residents who chose an option. eGFR in mL/min/1.73 m2 may be converted to mL/s/1.73 m2 by multiplying by 0.01667; serum creatinine in mg/dL may be converted to μmol/L by multiplying by 88.4. Abbreviations: ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker; BP, blood pressure; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate; PGY, postgraduate year.

Total Score of Internal Medicine Residents by PGY (maximum score = 30) 

The mean score of all PGYs was 21.5 ± 4.3 (71.8% ± 14.4%). The mean score of PGY1 was 20.6 ± 4.6 (68.8% ± 15.4%), mean score of PGY2 was 21.8 ± 4.4 (72.9% ± 14.7%), and mean score of PGY3 was 22.2 ± 3.6 (74.0% ± 12.0%; P = 0.004; Fig 1). Regression analysis was performed to evaluate the final performance score as predicted by PGY, sex, type of program (university versus community), and international training. These 4 variables accounted for a variance of r2 = 0.026 (P = 0.02). PGY was the only significant predictor of the final performance score in the survey (P = 0.002), and with each increase in PGY, the final performance score increased by 0.82 (2.7%). PGY also remained significant (P = 0.005) in the random intercept model, with each increase in PGY associated with an increase in final score by 2.5%.

View Large Image Download to PowerPoint

Figure 1. Awareness and knowledge of Kidney Disease Outcomes Quality Initiative clinical practice guidelines in internal medicine residents according to postgraduate year (PGY). Light blue bars, PGY1; maroon bars, PGY2; and light yellow bars, PGY3. Mean score in each section is reported as percentage of total score for that section. *Significant improvement in performance (P < 0.05) with increasing PGY was observed in definition of chronic kidney disease (CKD), classification of CKD, complications of CKD, nephrologist referral, and final performance.

Discussion 

Our study assessed the awareness of KDOQI guidelines for the management of CKD and identified significant gaps in this knowledge in a sample of internal medicine residents across the country. It also showed that as residents progressed in their residency year, their performance in certain areas of CKD management and overall performance improved. However, awareness of clinical practice guidelines for CKD was low across all PGYs, suggesting that incorporation of these guidelines into the medicine training curriculum is not robust at the present date.

The KDOQI guidelines were formulated by a combination of structured evidence review and expert opinion.1 Achieving clinical performance targets according to these guidelines is associated with decreased mortality, hospital admissions, and resource use in hemodialysis patients.7 The impact of these guidelines on patients in the early stages of CKD care requires further study. Multidisciplinary team–based care in a formal CKD clinic may improve survival in patients after they have started dialysis therapy.8 However, even in a dedicated CKD clinic, KDOQI goals are achieved in only a small proportion of patients.9

Primary care physicians are less likely to recognize progression of CKD and recommend nephrologist care. Lack of awareness of CKD clinical practice guidelines was identified as an important barrier to care.5 About one-third of residents were not aware of any guidelines for CKD, and another third believed the JNC-7 guidelines10 for hypertension and ADA guidelines11 for diabetes were sufficient to manage CKD. These findings are consistent with another study evaluating perceptions and practice patterns in CKD care among residents.12 Deficiencies also exist in resident knowledge of guidelines for cholesterol management (Third National Cholesterol Education Program),13 hypertension management (JNC-7),14 asthma management (National Heart, Lung, and Blood Institute),15 tuberculosis management (American Thoracic Society),16 and colorectal screening (American Cancer Society).17 The residents in our survey scored a mean 72%, which may be higher than for the other guidelines, possibly because some CKD guidelines were adapted from the JNC-7,18 ADA,19 and National Cholesterol Education Program.20

During internal medicine residency training, residents work with nephrologists mainly in inpatient consults. Thus, the only opportunity they may have to learn CKD management is in their medicine clinic. The internal medicine attending physicians themselves, like other primary care physicians, may be well aware of and using guidelines for hypertension and diabetes, but may be less aware of CKD guidelines.6 A few reasons for this pattern of care can be speculated. The guidelines for hypertension and diabetes are older, whereas the KDOQI guidelines are relatively new (February 2002) and some are still being developed and updated. The KDOQI guidelines are very extensive (>200 pages) and not concisely presented, like other guidelines. Clinical inertia, a common problem in physicians managing chronic diseases, may make it difficult to appreciate the effectiveness of the CKD guidelines.21 Also, CKD may be asymptomatic until later stages, when uremia sets in and causes nonspecific symptoms.

Our study identified gaps in the current knowledge of CKD guidelines in internal medicine residents. Half the residents were not aware that kidney injury (proteinuria) for 3 months or longer defines CKD. Another third believed that a patient needs to have decreased kidney function (GFR < 60 mL/min/1.73 m2) to have CKD. Also, a third of the residents were unable to identify CKD stage III. Knowledge of methods for early detection of CKD (with such markers of kidney damage as proteinuria) and CKD stages will allow an appropriate clinical action plan to be formulated.1 Residents were less aware of the following CKD risk factors: elderly age, obesity, African American/Hispanic race, systemic lupus erythematosus, nonsteroidal anti-inflammatory drug use, and family history of CKD.2 Fewer residents were aware of such relevant laboratory measurements as urinalysis, urine dipstick, and random urine albumin or protein-creatinine ratio to assess kidney damage.2 Fewer residents were aware of such dietary and lifestyle changes as salt restriction, lipid control, weight loss, and smoking cessation to reduce the progression of CKD, although there is insufficient evidence for this recommendation.2 Very few residents recognized nondihydropyridine calcium channel blockers as additive antihypertensive agents to decrease proteinuria.22 Only half the residents were aware of CKD putting a patient in the “highest risk” group for subsequent cardiovascular events.23 They were also less aware of other CKD complications, including malnutrition, neuropathy, and reduced functioning and well-being.

There were many areas in which the residents matched their expected performance. All residents identified the traditional risk factors for CKD: diabetes and hypertension. Most residents were aware of using estimated GFR to assess renal function, rather than using serum creatinine level alone. Most residents knew the goal BP for patients with CKD. They performed very well in their clinical action plan to slow the progression of CKD by means of BP control using ACE inhibitors/ARBs and aggressive glycemic control. Almost all residents identified ACE inhibitors/ARBs in decreasing proteinuria. Most residents identified anemia and bone disorder as potential complications of CKD. Finally, most residents chose to refer a patient with GFR less than 30 mL/min/1.73 m2 to a nephrologist.

Our study has important implications for program directors of internal medicine residency programs and nephrologists working with residents. Residency training may offer the best opportunity to improve the awareness and knowledge of CKD guidelines through more focused educational efforts, such as noon conference, grand rounds, and lectures by nephrologists. The impact of these traditional learning methods on resident learning have not been very encouraging.24 Problem-based learning25 or case-based learning26 may be more effective in improving knowledge. Medicine teaching faculty can encourage residents to identify patients at high risk of CKD and perform appropriate evaluation and management. Nephrology elective rotations may incorporate resident learning in a CKD clinic. In a program such as ours that is starting a CKD clinic, attending nephrologists can use this information to improve the performance of the rotating residents. The NKF may need to design concise educational materials, such as pocket cards and compact discs, to impart knowledge of the CKD guidelines without overwhelming the busy physician. Such efforts during training potentially could have a positive impact on the management of CKD as residents graduate to become independent physicians.

Our study also shows that performance among residents improved by the PGY, which attests that clinical experience and learning help improve residents' knowledge. The difference in scores among the PGYs was small, although significantly different, likely because of the small SD in scores. Additional studies need to evaluate whether educational efforts to improve knowledge in residents translate into better clinical outcomes. Barriers to learning and implementation of the CKD guidelines should be identified and interventions should be carried out to overcome these barriers.

Our study has several strengths. The multi-institutional random sample of residents with almost equal distribution from university and community programs improved the generalizability of our findings to other institutions. We designed the survey from the official KDOQI guidelines, thus avoiding differences in opinion to a large extent. The online nature of the survey allowed respondents to complete the survey quickly. This is evident from the number of responses we received during the short study period. Also, the questionnaire instrument had a modest Cronbach α reliability index that we consider adequate for an instrument testing knowledge in various themes of CKD management (questionnaire analysis under review elsewhere).

We also recognize several limitations to our study. We do not have data for nonresponders, and this bias could not be accounted for. Respondents self-selected to answer the questionnaire. We kept the questionnaire short to ensure maximum responses while covering many aspects of CKD care. Thus, there were very few questions testing each theme of CKD management, and this may less accurately evaluate the level of knowledge. Our study was conducted during a short period and we did not send reminders that may have further increased the number of responses. Most residents were from programs in the Northeast and Midwest. Although there are links between theoretical knowledge and practical application, our study could not identify whether the residents behave in the manner consistent with the reported answers.

In conclusion, internal medicine residents have significant gaps in their knowledge of CKD clinical practice guidelines. Educational efforts during residency training are needed to increase awareness of these guidelines, which may improve CKD management and clinical outcomes.